arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2602.10312cs.LG
收藏

无需训练的检索增强生成与强化推理用于洪水损害现在预测

Training-free retrieval-augmented generation with reinforced reasoning for flood damage nowcasting

Urban Resilience.AI Lab, Zachry Department of Civil and Environmental Engineering(城市韧性.AI实验室,土木与环境工程系);Department of Computer Science and Engineering(计算机科学与工程系);Department of Civil, Environmental and Architectural Engineering(土木、环境与建筑工程系);Institute for a Disaster Resilient Texas(德克萨斯灾害韧性研究所)

Lipai Huang, Kai Yin, Chia-Fu Liu, Ali Mostafavi

AI总结:

本文提出R2RAG-Flood框架,通过强化推理实现无需训练的洪水损害现在预测,利用结构化预测器、文本摘要和推理轨迹构建知识库,在推理阶段通过地理邻近和自由样本支持案例推理,提升预测准确性与成本效率。

摘要:

我们提出R2RAG-Flood,一种无需训练的检索增强生成框架,用于洪水损害现在预测中的强化推理。该框架从带标签的表格记录中构建以推理为中心的知识库,每个样本包含结构化预测器、紧凑的文本摘要和模型生成的推理轨迹。在推理过程中,目标提示通过地理局部邻居和选择的自由样本增强,支持基于案例的推理,而无需任务特定的微调。两阶段流程首先确定损害发生,然后在三个级别的财产损害范围(PDE)分类中细化严重性,随后进行保守降级检查以验证弱支持的过度严重输出。在得克萨斯州哈里斯县的哈维飓风案例研究中,监督表格基线达到0.714的整体准确率和0.859的受损类别(中等和高PDE)准确率。在七个LLM骨干网络上,R2RAG-Flood达到0.613-0.668的整体准确率和0.757-0.896的受损类别准确率,同时为每个预测提供结构化理由。在本研究使用的严重性-成本指标下,较轻的R2RAG-Flood变体比监督基线和较大的LLM骨干网络更具成本效益。这些结果展示了以推理为中心、无需训练的洪水损害现在预测在现实案例研究环境中的可行性。

英文摘要:

We propose R2RAG-Flood, a training-free retrieval-augmented generation framework for flood damage nowcasting with reinforced reasoning. The framework builds a reasoning-centric knowledge base from labeled tabular records, where each sample includes structured predictors, a compact text-mode summary, and a model-generated reasoning trajectory. During inference, the target prompt is augmented with geographically local neighbors and selected free-shots to support case-based reasoning without task-specific fine-tuning. A two-stage procedure first determines damage occurrence and then refines severity within a three-level Property Damage Extent (PDE) classification, followed by a conservative downgrade check for weakly supported over-severe outputs. In a Hurricane Harvey case study in Harris County, Texas, the supervised tabular baseline achieves 0.714 overall accuracy and 0.859 accuracy on the damaged classes (medium and high PDE). Across seven LLM backbones, R2RAG-Flood achieves 0.613--0.668 overall accuracy and 0.757--0.896 accuracy on the damaged classes while providing a structured rationale for each prediction. Under the severity-per-cost metric used in this study, lighter R2RAG-Flood variants are more cost-efficient than the supervised baseline and larger LLM backbones. These results demonstrate the feasibility of a reasoning-centric, training-free pipeline for flood damage nowcasting in a realistic case-study setting.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑